Improve database seeding

Two new models, namely likes and follows have been added to the database schema.
Seed data should be written for these as well.
Would be nice if the posts had different amount of likes and not everyone should follow everyone.

Start by locating the database schema definitions for likes and follows and the existing seed-data entry point. Add seed records for both relationships, ensuring posts receive varied like counts and users do not all follow one another; verify the seeded database reflects those constraints.

Written by the indexing model from the issue text.